Earlier this week I wondered about how to engineer a 'negotiated withdrawn' of aid to avoid fast-growing countries getting trapped in aid dependency.
Then I came across an interesting article in the Financial Times by Adrian Wood, Chief Economist of DFID. Wood argues that we should limit aid to a certain proportion of a country's budget - say 50%, or maybe 10% of GDP. Bill Easterly and Robert Wade provide trenchant commentary. (Nothing new here, says Easterly, but it won't work - the incentives for donors are to continue putting out aid come what may).
The debate continues at the Center for Global Development, with contributions from Nancy Birdsall, Jeff Sachs and Michael Lipton, amongst others. (Surely the problem is not too much aid, but too little? says Sachs - particularly when we have promised it and then not delivered, as is happening now).
I'm all for setting a limit to aid: but please let's make it a time limit, not a quantity limit. As Jeff Sachs points out, 10% of GDP for a country with a GDP per capita of $200 is $20 per person per year. That might be the upper limit of what a capacity-strapped or corrupt government can spend, but in post-war or desperately poor countries, much more will have to be directed at (re)building infrastructure, if necessary bypassing the government. To give an example off the top of my head, Liberia's annual budget is about $200m (itself the highest for 15 years), rebuilding their old hydropower station would cost at least $200m. I'd be curious to know what proportion of German or Japanese GDP was spent on rebuilding in the period 1945-50.
Rather than limiting expenditure per year, I'd like to see an aid agenda that says "We will work with you to achieve these targets and build capacity - but after 2015 we will begin cutting aid - and by 2025 we will have shut up shop, sold our Land Cruisers and our country economists will be out of a job. Over to you." Call it a surge, then a staged withdrawal.